## Warning: Too few values at 454677 locations: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
## 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, ...

Federal Laboratory Funding in the 21st Century - A 15 Year Review

Introduction

Methods

Data Collection

NIH budget files for fiscal years 2000 through 2014 were collected from the NIH RePORTER website’s export tool, ExPORTER (website citation). Fiscal year, NIH Application ID, Primary Investigator ID Number and Total Cost data was collected for each fiscal year. Primary Investigator ID numbers were used to keep all data anonomyous. Data for all R01-equivelent awards were collected, which include the activities R01, R23, R29, and R37, although not all awards may have been used in a given year. The breakdown of which awards were used in each year can be seen in table 1 (MAKE TABLE 1). Total cost information was associated with each investigator for each fiscal year.

Starting in 2007, R01-equivelent awards could have more than one primary investigator. To account for this, the total cost dispersed per application was divided by the number of primary investigators and this number was associated with each individual investigator. This operates on the assumption that award funds are being dispersed evenly to each investigator, but we realize in practice this may not be the case.

Per year, the number of grants per investigator was calculated. This was a count of how many unique Application IDs were associated with each investigator ID. The sum cost was calculated as the sum of the total cost associated with that investigator for each fiscal year.

Year 2000 adjusted dollars were calculated using the Consumer Price Index from the Bureau of Labor Statistics (CITATION - BLS link). An adjustment factor was calculated and applied to the total cost amounts for each year after year 2000.

Results

Gini coefficient for inequality

The Gini coefficient has been used in previous literature to characterize inieuality in income distributions (CITATION). The coefficient ranges from 0, where income is perfectly distributed among all group members, to 1, where income is perfectly concentrated with one group member. A larger coefficient signifies a greater income disparity. We used the Gini coefficient to describe R01-equivelent award disparity for each fiscal year, and it can be seen in Figure 1.

Discussion